View this post on the web at [link removed]

By Joshua Levine [ [link removed] ]
Cloud computing is at the center of digital commerce. Cloud services enable firms to migrate data and access it through “the cloud,” making information available anywhere and anytime. This allows [ [link removed] ] both startups and incumbents to reduce upfront costs and access bespoke services in addition to on-demand compute, a key input for AI model development and deployment. The market for cloud services is growing steadily, with three firms at the forefront: Amazon, Google and Microsoft. These firms are colloquially referred to as hyperscalers [ [link removed] ], a term that denotes their presence across the layers of technology and infrastructure that support cloud computing operations.
But cloud computing is under attack. Regulators in Europe, the U.K. and the U.S. have all opened investigations into such firms, insinuating that antitrust action or new regulation may be warranted. While competition enforcers play an important consumer protection role in various industries, including cloud computing, imposing new regulations that would limit cloud computing providers’ ability to grow and innovate is the wrong approach. A regulatory posture that encourages investment, allows for differentiation and customization and keeps costs low would further promote competition and choice in this vital market.
What Is the Cloud?
Before the advent of the cloud, if a firm wanted to store data, it had to purchase servers and manage its own information infrastructure. While this gave the firm complete control over its data, it required renting space in a data center or building infrastructure internally, a significant capital investment. By contrast, cloud computing services can provide that infrastructure for firms, along with other products and services, which lowers those firms’ operating costs.
Cloud computing services are divided into three categories: infrastructure (IaaS), platform (PaaS) and software (SaaS). The infrastructure layer is the most basic part of the cloud, as it stores data and provides direct-to-user computing resources. The platform layer [ [link removed] ] provides users with specific tools to manage and manipulate the data housed within the cloud. Finally, the software layer is the software products and services that can be added to or plugged into the platform, such as Salesforce or Slack. This layer has seen its growth explode [ [link removed] ] since 2015, and that growth continues to trend upward.
The market for cloud computing has similarly seen consistent [ [link removed] ] growth and investment over the past decade, but a few firms stand out. Amazon, Google and Microsoft, the hyperscalers, currently constitute [ [link removed] ] 65% of the cloud computing market in the U.S. The cloud computing market’s Herfindahl-Hirschman Index [ [link removed] ] (HHI) score, a metric used to measure market concentration, is 1,542. When the HHI analysis is expanded to include firms’ market shares that are at or below 1%, the HHI increases to 1,570. Based on the federal government’s current [ [link removed] ] antitrust merger guidelines (which count an HHI score above 1800 as highly concentrated), this market would be considered “moderately” concentrated, meaning merger activity by a hyperscaler would likely be scrutinized, particularly if Amazon (with 31% market share) is the acquiring party.
Clouds of Competition
The success of the hyperscalers, recent cloud price increases and growing concerns related to vendor lock-in—when technical and economic costs make changing cloud providers prohibitively expensive—have tripped the alarms of competition authorities around the world.
Last June, for example, the EU’s competition [ [link removed] ] authorities opened an investigation and sought charges against Microsoft’s Azure cloud for what it deemed restrictive licensing practices and high egress fees—that is, fees charged by cloud providers to migrate data to another cloud. Microsoft later agreed to pay [ [link removed] ] a fine of 20 million euros and develop a software product that allows alternative European cloud providers to run Microsoft Teams, as well as other Microsoft software. Microsoft also now lets [ [link removed] ] users license their software on other cloud providers.
Ofcom, the U.K.’s telecommunications regulator, published a report [ [link removed] ] highlighting egress fees and increasingly advanced offerings from incumbents as potential sources of consumer lock-in, before referring the case to the Competition and Markets Authority for a formal inquiry. The report notes, however, that individuals and firms interviewed as part of the report were not terribly concerned about any of these areas and few requested any public action. Rather, a wait-and-see approach was preferable to any new regulation of the public cloud market.
Last but not least, the U.S. Federal Trade Commission (FTC) released [ [link removed] ] an investigative report on partnerships between cloud service providers and AI developers. The study highlighted arrangements that provided discounts and contractual incentives for developers to use a single cloud provider rather than spread their spending across multiple clouds. FTC Commissioners Andrew Ferguson [ [link removed] ] and Melissa Holyoak [ [link removed] ], however, dissented to part of the release of the report because of its limited scope and speculation about future areas of concern or anticompetitive conduct. Specifically, the commissioners took issue with the report’s implication that the market already presented anticompetitive behavior, and that any further action would begin with this framing.
While concerns about competition in a market with “moderate” concentration are understandable, attempts to impose neutrality mandates [ [link removed] ] or prevent [ [link removed] ] vertically integrated offerings would ignore the benefits such services provide for businesses, oversimplify the cloud market and fail to consider alternative approaches that could spur greater choice and flexibility for cloud customers.
Looking Through the Clouds
The interest in AI development and deployment, along with an increasing demand for compute capacity generally, have buoyed overall demand for cloud computing services. As a result, prices for accessing cloud services have increased [ [link removed] ]. But this is a result of a sustained demand for such services, as market reports have documented [ [link removed] ]. While cloud providers are all providing the same basic service, hyperscalers and new entrants are pursuing different paths to compete and wrest market share from one another.
For hyperscalers, while they do emphasize price, they also provide a host of PaaS and SaaS components to support their individual cloud ecosystem. A prime example is Google [ [link removed] ] removing egress fees and Amazon [ [link removed] ] and Microsoft [ [link removed] ] following suit, in order to remain competitive. Not wanting to be caught with a more expensive service or to attract regulatory ire, the two market leaders followed [ [link removed] ] the lead of the third-place competitor.
Smaller providers are focusing on specific customer needs and new avenues to differentiate their products and undercut the hyperscalers on price and flexibility. HP [ [link removed] ]’s GreenLake hybrid offerings, for example, are aimed directly at enterprise customers rather than individuals, looking to service the business-to-business market. IBM [ [link removed] ]’s approach is similar, and it has adopted portability standards to reduce friction for businesses taking advantage of features from multiple providers. Oracle [ [link removed] ] highlights its own in-house offerings that are priced below those of the hyperscalers as a bid to regain its foothold within the market. And new AI-focused cloud provider CoreWeave [ [link removed] ] states that its per-hour cost beats that of competitors. Market research also points to businesses [ [link removed] ] adopting a multicloud strategy, with more than 80% of businesses taking advantage of multiple cloud providers.
If policymakers want to lower the barriers preventing new cloud providers from entering the market, they should focus on reducing costs with supply-side reforms. The largest cost factor for new data center construction is energy related: interconnection infrastructure. Long timelines [ [link removed] ] for grid interconnection and buildout within the United States, as well as the increasing energy demand for data centers (particularly those hosting [ [link removed] ] AI training and inference activities), create high barriers to entry for firms looking to expand cloud infrastructure. Accelerating construction timelines for energy infrastructure can be a game-changer for increasing data center construction and lowering a key input cost for the cloud computing services.
Regulators and competition enforcers, meanwhile, should focus on anticompetitive conduct, specifically that related to illegal tying, the practice of conditioning the purchase of one product with another, which violates U.S. antitrust law. Prior to launching regulatory actions against companies, absent obvious violations of antitrust law, the FTC and other competition enforcers should consider updating and expanding studies such as the one described above to better capture relevant aspects of the cloud market.
Beyond vigilance by regulators, incentivizing the adoption of standard protocols for interoperability and data portability is worth exploring. The Institute of Electrical and Electronics Engineers [ [link removed] ], the world’s largest professional technical association, released standards [ [link removed] ] in 2021 to support more interoperability between cloud computing providers. New entrants and smaller market participants have been early adopters of these standards in a bid to compete with established players. The National Institute of Standards and Technology already requires [ [link removed] ] government cloud providers to utilize portability and interoperability standards for federal uses when doing so is cost effective and secure. These rules could help smaller companies compete for contracts or serve as a use case on the effectiveness of open standards, potentially leading to more widespread adoption.
Furthermore, Congress should explore making it easier for new cloud service providers to compete for federal contracts. Last year, Congress considered legislation [ [link removed] ] that would require federal cloud contracts worth more than $50 million to adopt interoperable, multicloud strategies where “feasible and advantageous.” This legislation could create opportunities for new entrants to compete for federal contracts, particularly providers that focus on providing PaaS and SaaS capabilities designed particularly for the deployment of AI. Such moves would not be without tradeoffs, however, and may warrant further study to ensure that new requirements address tangible issues such as improving operational capabilities or creating savings for taxpayers, rather than simply undermining incumbents to score political points.
The cloud market is continuing to grow as the technology improves and unlocks gains for firms of all shapes and sizes. Concerns about oligopoly are understandable given the sheer size and power of the market leaders. But beyond the market-share shock value is a dynamic market defined by growing size, new entrants and increasing specialization. Rather than impose additional regulations or restrictions on existing entrants, legislation and regulation should focus on lowering barriers to building infrastructure and reducing friction and switching costs between providers. American firms are leaders in this key layer of the 21st-century technology stack, and we should ensure this status quo continues.
Joshua Levine is a research fellow at the Foundation for American Innovation.

Unsubscribe [link removed]?